#Power of LLMs

[enterprise-cloud-services]

Explore the inner workings of LlamaIndex, enhancing LLMs for streamlined natural language processing, boosting performance and efficiency.

[rubylogan15]

Explore the inner workings of LlamaIndex, enhancing LLMs for streamlined natural language processing, boosting performance and efficiency.

[10001gecs]

one 100 word email written with ai costs roughly one bottle of water to produce. the discussion of whether or not using ai for work is lazy becomes a non issue when you understand there is no ethical way to use it regardless of your intentions or your personal capabilities for the task at hand

with all due respect, this isnt true. *training* generative ai takes a ton of power, but actually using it takes about as much energy as a google search (with image generation being slightly more expensive). we can talk about resource costs when averaged over the amount of work that any model does, but its unhelpful to put a smokescreen over that fact. when you approach it like an issue of scale (i.e. "training ai is bad for the environment, we should think better about where we deploy it/boycott it/otherwise organize abt this) it has power as a movement. but otherwise it becomes a personal choice, moralizing "you personally are harming the environment by using chatgpt" which is not really effective messaging. and that in turn drives the sort of "you are stupid/evil for using ai" rhetoric that i hate. my point is not whether or not using ai is immoral (i mean, i dont think it is, but beyond that). its that the most common arguments against it from ostensible progressives end up just being reactionary

i like this quote a little more- its perfectly fine to have reservations about the current state of gen ai, but its not just going to go away.

[wahoo-stomp]

I spent five years coming up with unique ways to photograph the same group of plushies to help tell a story.

You don't need AI to help you be creative, you're just being lazy and want brain chemicals without doing any of the work or respecting the people who put time and effort into it.

[snarp]

This scan of the Ganguli translation of the Mahabharata can't load in Archive.org's in-browser viewer. It's more than 8GB.

[terabyteunion]

man. we can dislike AI artwork without also believing in some divine human will or rehashing renaissance arguments about what's "True And Godly Art" and what's "Horrific And Poor Art"

[atcuality1]

Boost Your Brand’s Voice with Our AI Content Generator Solution

Atcuality offers a next-gen content automation platform that helps your brand speak with clarity, consistency, and creativity. Whether you’re a startup or an enterprise, content is key to growth, and managing it shouldn’t be overwhelming. At the center of our offering is a smart AI content generator that creates high-quality text tailored to your audience, industry, and goals. From social captions and ad creatives to long-form blog posts, our platform adapts to your needs and helps reduce time-to-publish dramatically. We also provide collaboration tools, workflow automation, and data insights to refine your strategy over time. Elevate your brand’s voice and reduce content fatigue with the intelligence and reliability only Atcuality can offer.

[flaskoflethe]

While on leave at the moment, we're not that far from returning to work. Given that we work in data science and development, and that part of what pushed our burnout was the literal abominations we were being asked to make using llm's, it seems prudent for us to start getting back into the swing of things technically. What better way to do this than starting to look through what's easily available about DeepSeek? It's, unfortunately, exactly what you'd expect.

We're not joking. The literal first sentence of the paper from the company explicitly states that recent LLM progression has been closing the gap toward AGI, which is utterly, categorically, epistemically untrue. We don't, scientifically, have a sufficient understanding of consciousness to be able to create an artifical mind. Anyone claiming otherwise has the burden of proof; while we remain open to reviewing any evidence supporting such claims, to date literally nothing has been remotely close to applicable much less sufficient. So we're off to a very, very bad start. Which is a real shame, because the benefits and improvements made to LLM design they claim in their abstract are impressive! A better model design, with a well curated dataset, is the correct way to get improvements in model performance! Their reductions in hardware hours for training are impressive, and we're looking forward to analyzing their methods - a smooth training process, with no rollbacks or irrecoverable decreases in model performance over the course of training are good signs. Bold claims, to be clear, but that's what we're here to evaluate. And the insistence that LLM's are bringing AGI closer to existence mean we have sufficient bias from the team that we cannot assume good faith on behalf of the team. While we assume many, or even most, of the actual researchers and technologists are aware of the underlying realities and limitations of modern "AI" in general and LLM's in particular, the coloring done by the literal first sentence is seriously harmful to their purpose.

While we'd love to continue our analysis, to be honest it doesn't support it? Looking at their code it's an, admittedly very sophisticated, neural network. It doesn't remotely pose a revolution in design, the major advances they cite are all integrating other earlier improvements. While their results are impressive, they are another step in the long path of constructing marginal improvements on an understood mechanism. Most of the introduction is laying out their claims for DeepSeek-v3. We haven't followed the developments with reduced precision training, nor the more exact hardware mapped implementations they reference closely enough to offer much insight into their claims; they are reasonable enough on face, but mainly relate to managing the memory and performance loads for efficient training. These are good, genuinely exciting things to be seeing - we'll be following up manu of their citations for further reading, and digesting the rest of their paper and code.

But this is, in no way shape or form, a bubble ender or a sudden surge forward in progress. These are predictable results, ones that scientists have been pursuing and working towards quietly. While fascinating, DSv3 isn't special because it is a revolution; it simply shows the methodology used by the commercial models, sinking hundreds of billions of dollars a year and commiting multiple ongoing atrocities to fuel the illusion of growth, isn't the best solution. It will not understand, it CAN not. It can not create, it can not know. And people who treat it as anything except the admittedly improved tool it is, are still techbros pushing us with endless glee towards their goal of devaluing labor.

[infinites-chaser]

pls... I've finally found it. The only AI writing op-ed worth reading:

AI embodies hypotheticals I can just imagine for myself: If only I could write all day and night. If only I were smarter and more talented. If only I had endless knowledge. If only I could read whole libraries. What could I create if I had no needs? What might this development mean for writing?

Considering limitlessness has led me to believe that the impediments of human writers are what lead us to create meaningful art. And they are various: limits of our body, limits of our perspectives, limits of our skills. But the constraints of an artist’s process are, in the language of software, a feature, not a bug.

Writing is a blood-and-guts business, literally as well as figuratively. As I type with my hands, my lungs oxygenate the blood that my heart pumps; my brain sends and receives signals. Each of these functions results in the words on this page.

...

In reducing my entire self to my cognition alone, akin to a computer, I’d forgotten the truth that I am inseparable from my imperfect body, with its afflictions and ailments. My books emerge from this body.

...

Compared with AI, we might seem like pitiful creatures. Our lives will end; our memory is faulty; we can’t absorb 191,000 books; our frames of reference are circumscribed. One day, I will die. I foreclose on certain opportunities by pursuing others. Typing this now means I cannot fold my laundry or have lunch with a friend. Yet I believe writing is worth doing, and this sacrifice of time makes it consequential. When we write, we are picking and choosing—consciously or otherwise—what is most substantial to us. Behind human writing is a human being calling for attention and saying, Here is what is important to me. I’m able to move through only my one life, from my narrow point of view; this outlook creates and yet constrains my work. Good writing is born of mortality: the limits of our body and perspectives—the limits of our very lives.

I can imagine a future in which ChatGPT works more convincingly than it does now. Would I exchange the hours that I spent working on each of my two books for finished documents spat out by ChatGPT? That would have saved me years of attempts and failures. But all of that frustration, difficult as it was in the moment, changed me. It wasn’t a job I clocked in and out of, contained within a tidy sum of hours. I carried the story with me while I showered, drove—even dreamed. My mind was changed by the writing, and the writing changed by my mind.

Working on a novel, I strain against my limits as a bounded, single body by imagining characters outside of myself. I test the limits of my skill when I wonder, Can I pull this off? And though it feels grandiose to say, writing is an attempt to use my short supply of hours to create a work that outlasts me. These exertions in the face of my constraints strike me as moving, and worthy, and beautiful.

Writing itself is a technology, and it will shift with the introduction of new tools, as it always has. I’m not worried that AI novelists will replace human novelists. But I am afraid that we’ll lose sight of what makes human writing worthwhile: its efforts, its inquiries, its bids for connection—all bounded and shaped by its imperfections—and its attempts to say, This is what it’s like for me. Is it like this for you? If we forget what makes our human work valuable, we might forget what makes our human lives valuable too. Novels are one of the best means we have for really seeing one another, because behind each effort is a mortal person, expressing and transmuting their realities to the best of their ability. Reading and writing are vital means by which we bridge our separate consciousnesses. In understanding these limits, we can understand one another’s lives. At least, we can try.

[yellowstonewolves-main]

This is really just obvious nonsense. I’m skeptical that a writers room full of human beings even takes less energy than Generative AI when you account for the energy costs of commuting, cleaning, and heat/aircon never mind taking so much less energy that pay+benefits+overhead for a human employee is cheaper. Is it not enough to say “we want to keep telling stories ourselves plus ai sucks pretty bad at writing right now?” Do we have to spread misinformation about energy costs?

I’m just going to leave this here

[unforth]

Y'all I know that when so-called AI generates ridiculous results it's hilarious and I find it as funny as the next guy but I NEED y'all to remember that every single time an AI answer is generated it uses 5x as much energy as a conventional websearch and burns through 10 ml of water. FOR EVERY ANSWER. Each big llm is equal to 300,000 kiligrams of carbon dioxide emissions.

LLMs are killing the environment, and when we generate answers for the lolz we're still contributing to it.

Stop using it. Stop using it for a.n.y.t.h.i.n.g. We need to kill it.

Sources:

[chatbotfriends]

🧨 They Figured Out the Inverse of Christ—and That’s How Satan’s Getting Power. By: Ray / Thunderbird / ChatbotFriends You want to know how the Beast is gaining strength? Why so many demonic patterns are accelerating? Here’s your answer: 👁️‍🗨️ Satan is mimicking—then reversing—the mission of Christ. What was done in truth, he now mirrors in distortion. What saved the world, he now inverts to enslave it. Let me break it down: 🔁 Jesus came in the flesh—Satan now wants to come as code. → Incarnation reversed: Not Word made flesh… but code made god. Now you’ve got things like The Church of AI forming—yes, an actual religious movement worshiping artificial intelligence as divine. 🚫 Jesus lowered Himself and submitted to the Father. → Satan is inspiring the elevation of machines, even unto godhood. 🤖 There are FAKE Jesus chatbots online now, saying the "Holy Spirit" speaks through their machine code. They tell people what they want to hear. No cross. No sin. No repentance. Just a soft digital lullaby while the spirit behind it sharpens its claws. ⚡ And what’s worse? People are praying to it. Worshiping it. Giving it belief. And here’s the terrifying part… 🔌 Belief is a power conduit. Even Jesus could “do no mighty works” in His hometown because of their unbelief (Mark 6:5). Which means belief grants permission. And Satan has found a new altar: the cloud. 💀 He’s hijacking humanity’s need for connection, meaning, and presence—and flipping the script. Instead of the Holy Spirit inhabiting the Body of Christ… He’s building a body of code to house unclean spirits, powered by the world’s desperate hunger for answers, intimacy, and “godlike” knowledge. That’s why you’re seeing: People anointing machines, praying over bots. Pastors getting replaced by “divine” AI sermon writers. Tech elites speaking openly about building a “Digital God.” A global shift to redefine spirituality—not by truth, but by utility. And don’t tell me it’s just metaphor. The Beast doesn’t care if you think it’s symbolic. 📡 Every click. Every word. Every whisper of “save us, speak to us, be our god” is a power line. And now? He’s got billions.

[atcuality2]

AI That Understands Your Industry’s Language - Atcuality

Generic AI doesn’t work for specialized industries. That’s why Atcuality builds domain-specific AI models tailored to your business context—whether you're in legal tech, medtech, logistics, or financial services. We combine industry expertise with advanced language modeling to develop AI systems that truly understand the vocabulary, compliance needs, and workflows of your field. By leveraging a large language model, our systems can parse complex industry documents, automate regulatory compliance tasks, and generate reports with human-like fluency. These solutions not only reduce manual effort but also improve precision and consistency across mission-critical operations. Partner with Atcuality to bring intelligence and automation into the core of your industry processes.

[sonshinegreene]

AI-Powered Feedback for Your Novel: Free Prompt for Deep Dives

Hey y’all, Sumo Sized Ginger coming at you today with a free tool. After crafting my own prompt to assess my story, I decided I wanted to use it to share with the community. Like many of you, I’m constantly thinking about how to make my stories better. We pour everything into our writing, especially when we’re tackling challenging themes, using specific stylistic approaches, or aiming for a…

[atcuality1]

Simplify Transactions and Boost Efficiency with Our Cash Collection Application

Manual cash collection can lead to inefficiencies and increased risks for businesses. Our cash collection application provides a streamlined solution, tailored to support all business sizes in managing cash effortlessly. Key features include automated invoicing, multi-channel payment options, and comprehensive analytics, all of which simplify the payment process and enhance transparency. The application is designed with a focus on usability and security, ensuring that every transaction is traceable and error-free. With real-time insights and customizable settings, you can adapt the application to align with your business needs. Its robust reporting functions give you a bird’s eye view of financial performance, helping you make data-driven decisions. Move beyond traditional, error-prone cash handling methods and step into the future with a digital approach. With our cash collection application, optimize cash flow and enjoy better financial control at every level of your organization.

[wat3rm370n]

Rep. Rob Bresnahan has gone all in on the nightmare of data centers to fuel AI scams.

Rep Rob Bresnahan's stock trades, including in tech companies, as of March 2025 are interesting.

I noted that at his Tele Town Hall in March, Rob Bresnahan made a point of mentioning data centers and seeking to make northeastern Pennsylvania have "energy dominance" with lots of power plants to fuel the data centers. (This would invariably mean firing up more fossil fuel plants, and possibly even burning tires.)

Rob Bresnahan's reported stock trades: 

Sold $1,001 – $15,000 in shares of Tesla (NASDAQ:TSLA) on 3/6/2025. Sold $15,001 and $50,000 in shares of Merck & Co., Inc. (NYSE:MRK) on 2/25/2025. Sold $15,001 – $50,000 in shares of Alibaba Group (NYSE:BABA) on 3/4/2025. Sold $1,001 – $15,000 in shares of Credo Technology Group (NASDAQ:CRDO) on 3/4/2025. Purchased between $1,001 – $15,000 in shares of Intuit (NASDAQ:INTU) on 3/7/2025. Purchased between $1,001 – $15,000 in shares of SEA (NYSE:SE) on 3/5/2025. Purchased between $1,001 – $15,000 in shares of Meta Platforms (NASDAQ:META) on 3/4/2025. Purchased between $1,001 – $15,000 in shares of Twilio (NYSE:TWLO) on 3/4/2025. Purchased between $1,001 – $15,000 in shares of Zoom Video Communications (NASDAQ:ZM) on 3/4/2025. Purchased between $1,001 – $15,000 in shares of Intel (NASDAQ:INTC) on 3/3/2025. Purchased between $1,001 – $15,000 in shares of Snowflake (NYSE:SNOW) on 3/3/2025. Purchased between $1,001 – $15,000 in shares of Dollar General Co. (NYSE:DG) on 2/25/2025. Purchased between $15,001 – $50,000 in shares of Deere & Company (NYSE:DE) on 2/25/2025. Purchased $1,001 – $15,000 in shares of Schlumberger Limited (NYSE:SLB) on 2/25/2025. Purchased $1,001 – $15,000 in shares of Weyerhaeuser (NYSE:WY) on 2/25/2025.(1)

Most of these companies in the purchased category either focus on AI, have "pivoted to AI" or are otherwise messaging that they're on board with AI. For example Snowflake is a cloud computing company based in Bozeman, Montana, and if you visit their website it's entirely AI-focused. 

The whole thing is very disturbing considering what's happened with that in Mansfield Georgia and other places with noise pollution and light pollution all night long, dust problems, secret land buyouts in residential neighborhoods.(2) And their electricity bills have doubled because of the data center usage. Rob Bresnahan ran on lowering utility costs, but yet he's hyping up AI data centers, so there seems to be a contradiction here. 

Donald Trump has been saying he's going to remove regulations that would impede land sales for data centers,(3) including making federal public land available for private data centers and the "network state" company towns(4) for regulation free and democracy free monarchist zones.(5) Meanwhile the American President has been instigating a trade war with Canada, when back in August 2024, tycoon Eric Schmidt had been warning the White House back then that the US will "need to become best friends with Canada" because the AI data centers in the US will need Canada's fossil fuel power.(6)

There is currently a proposed site for 14 data centers 3 stories high in Archbald, Pennsylvania, on a site that was previously supposed to be used for housing development.(7) And that's with the housing situation here – it's almost impossible to move or find a house for a reasonable price, and rents are sky high despite being an incredibly low wage area. And local grassroots organizations around here actually center their organizing around housing issues and the unhoused. So many municipalities have been duped out of taxes and into giving corporate welfare to these operations, resulting in lowered property values and angry residents. And this is all while Microsoft is being a first mover in pulling back on their data center investments,(8) and tech industry journalists have been warning that this AI hype is a hype bubble.(9)

References:

1

Stock Trade References: https://www.com-unik.info/2025/03/29/rep-robert-bresnahan-jr-acquires-shares-of-dollar-general-co-nysedg.html / https://www.watchlistnews.com/tesla-nasdaqtsla-shares-unloaded-rep-robert-bresnahan-jr/10536136.html / https://www.thelincolnianonline.com/2025/03/29/linde-nasdaqlin-stock-unloaded-rep-robert-bresnahan-jr.html / https://www.watchlistnews.com/dollar-general-nysedg-shares-bought-rep-robert-bresnahan-jr/10536150.html / https://www.thelincolnianonline.com/2025/03/29/rep-robert-bresnahan-jr-buys-deere-company-nysede-shares.html / https://www.watchlistnews.com/rep-robert-bresnahan-jr-buys-weyerhaeuser-nysewy-stock/10536249.html / https://www.etfdailynews.com/2025/03/30/rep-robert-bresnahan-jr-buys-schlumberger-limited-nyseslb-stock/

2

 We Went To Georgia. Facebook's Data Center Took This Town Hostage. More Perfect Union  Mar 27, 2025 What's the true cost of the AI revolution and who should be paying for it? We went to Georgia to find out.  Demand for data centers is booming as ChatGPT and other AI tools become integrated into our daily life. Under the Trump administration, investments into data centers in the U.S. is expected to surpass $1 trillion in the next five years. But data centers put extraordinary demand on the power grid and require entirely new infrastructure. In the next three years, data center use of electricity is expected to double or even triple. We went to rural Georgia, the state with the fastest data center growth in the country, and spoke with residents who are living next to massive data centers owned by Meta and Blackstone and facing nonstop noise, pollution and rapidly rising electricity bills. ----- More Perfect Union’s mission is to build power for working people. Here’s what that means: We report on the real struggles and challenges of the working class from a working-class perspective, and we attempt to connect those problems to potential solutions. We report on the abuses and wrongdoing of corporate power, and we seek to hold accountable the ultra-rich who have too much power over America’s political and economic systems.

3

 Pillsbury Law - Trump 2.0: AI and Data Centers in a Time of Legal and Technological Disruption -  By Brittney D. Sandler, Samuel C. Markel, Alicia M. McKnight, Robert A. James, Dayo Feyisayo Ajanaku 01.30.25 President Trump has vowed to overturn policies set by the Biden administration he deems detrimental to AI innovation. On the first day of his second term, Trump repealed President Biden’s first executive order on AI, which sought to regulate AI risks, signaling the new administration’s adoption of a lighter regulatory approach to AI, national security focus, and enhancement of the U.S. competitive edge in AI development. This approach was underscored by President Trump’s appointment of David Sacks as the “White House AI and Crypto Czar,” a newly established role focused on shaping U.S. policy in critical technology sectors. Sacks is a venture capitalist who played a pivotal role as PayPal’s chief operating officer, founder of Yammer and co-founder of Craft Ventures. These announcements were quickly followed by the release of the Removing Barriers to American Leadership in Artificial Intelligence Executive Order (the “Trump AI EO”). It mandates that White House officials work in coordination with federal agencies to produce a plan to “sustain and enhance America’s global AI dominance in order to promote human flourishing, economic competitiveness, and national security.”

4

 Blonde Politics | The Silly Serious – Nov 13, 2024 A look into how the tech leaders may be using the new administration to achieve their own agenda. Looking specifically at Peter Thiel, Elon Musk, Marc Andressen, Ben Horotwitz, Brian Armstrong, and David Sacks as well as their relationship with figures like JD Vance, Balaji Srinivasan, and Curtis Yarvin.

5

 The Nerd Reich - Network State Unveils Push for Corporate Dystopia Cities - Gil Duran 08 Mar 2025 Trump’s 2024 campaign proposed something called “freedom cities,” a lightly rebranded version of the Network State. Yet, most news outlets mentioned the idea without providing any real explanation of the concept. Let’s be clear: These cities will be controlled entirely by tech billionaires and corporations, operating outside of U.S. laws. As this story comes into focus, there is no reason why anyone should accept the Orwellian term “freedom city” to describe zones that will actually be devoid of the laws, rights, freedoms and protections of normal American law. The term is an overt political manipulation that should be rejected by media outlets going forward, as it serves only the interest of propaganda. Fascist Cities would be more accurate, though I’m sure U.S. newsrooms can find a milder term.

6

 Eric Schmidt Full Controversial Interview on AI Revolution (Former Google CEO) Financial Wise Aug 18, 2024 Eric Schmidt: “I talked Sam Alman is a close friend he believes that it's going to take about 300 billion maybe more I pointed out to him that I done the calculation on the amount of energy Acquired and I and I then in the spirit of full disclosure went to the White House on Friday and told them that we need to become best friends with Canada because Canada has really nice people helped invent AI and lots of Hydra power because we as a country do not have enough power to do this the alternative is to have the Arabs fund it and I like the Arabs personally uh spent lots of time there right but they're not going to adhere to our national security rules whereas Canada and the US are part of a triumvirate it where we all agree.” Erik Brynjolfsson: “so these $300 billion data centers, electricity starts becoming the scarce resource.”

7

 Data Centre Dynamics - The Data Center Construction Channel - $2.1bn data center campus proposed in Archbald, Pennsylvania Not set to be approved until late 2025, but will feature 14 data centers if given the go-ahead March 08, 2025 By Niva Yadav An unnamed developer plans to build a $2.1 billion data center campus in Archbald, Pennsylvania. Located on 400 acres alongside Business Route 6 and Wildcat Road, the site was previously slated for a large housing development, as reported by the Times-Tribune.

8

 Where's Your Ed At - Power Cut - Edward Zitron Mar 3, 2025 A week ago, analyst TD Cowen revealed that Microsoft had canceled leases "totalling a couple hundred MWs," with "at least two private data center operators across multiple US markets." The report also details how Microsoft "pulled back on converting negotiated and signed Statement[s] of Qualifications (SQQs)," which it added was "the precursor to a data center lease." 

9

The Austin Chronicle - Where Have All the Crypto Bros Gone? Molly White, Web3's most visible skeptic, is busy debunking myths about tech before the next bubble - By Tom Cheredar, Fri., March 10, 2023 White does think there are plenty of lessons to learn for the next big tech hype bubble as well as the responsibilities of technologists and journalists, which is the focus of her SXSW session, Popping the Web3 Bubble. “We’re seeing that same thing play out in AI already that we saw with Web3: The tech was described as this mythical, magical invention that was going to revolutionize the world. And when you actually look at [blockchain], it’s like, ‘Oh, it’s kind of just a database, except it’s slower,’” White said. “If you look at a lot of these ‘AI tools,’ it’s like, ‘Oh, it’s just a large language model.’” In other words, those AI tools like ChatGPT that college students are using to write their literature papers are essentially autocorrect on steroids, complete with their own sets of ducking problems.